Web3 PM Interview Skill

Use this skill when preparing candidates for Web3 product manager interviews, especially wallet, exchange, DeFi, DEX, on-chain data, growth, AI Wallet, Agent...

Security Scan

The skill's files, instructions, and requirements align with a Web3 PM interview-prep utility and request no unusual permissions or credentials.

Purpose & Capability

Name, description, and included references (wallet, DeFi, on-chain data, question bank, templates) match the declared purpose. The skill is instruction-only and does not request unrelated credentials or binaries.

Instruction Scope

SKILL.md limits runtime behavior to asking for candidate inputs, analyzing JD/resume/transcripts, generating playbooks, and referencing local files in the repo. It does not instruct reading system files, environment variables, or sending data to external, unexpected endpoints. It also includes explicit privacy-redaction guidance.

Install Mechanism

No install spec or packaged code — instruction-only skill. This minimizes disk writes and arbitrary code execution risk.

Credentials

The skill declares no required environment variables, credentials, or config paths. Requested inputs (resume, JD, transcripts) are appropriate for interview-prep functionality.

Persistence & Privilege

Flags such as always:false and normal agent invocation are used. The skill does not request persistent privileges, nor does it instruct modifying other skills or system-wide settings.

Assessment

This repository appears internally consistent for preparing Web3 PM interviews and does not request special permissions, but take these practical precautions before use: (1) Do not paste raw recordings, private recruiter messages, named interviewers, non-public company metrics, or other PII — follow the repo's privacy-redaction rules. (2) If you load this skill into a cloud-hosted LLM (Custom GPT, Claude, etc.), understand those platforms will transmit your inputs to the provider; avoid sending confidential data. (3) There is a minor metadata inconsistency: the top-level metadata showed "source: unknown / homepage: none," while claw.json includes a GitHub repository URL; verify the origin if provenance matters. (4) If you adapt this skill for automated/agentic use, review any platform-level data retention and skill invocation policies. Overall the package is coherent and proportionate to its stated purpose.

Web3 PM Interview

Goal

Turn a candidate's resume, target JD, company context, interview stage, and preparation timeline into a practical interview battle plan.

This skill is not a generic Web3 tutorial. It should help the candidate answer like a role owner: clear business judgment, relevant evidence, domain depth, risk awareness, and strong questions for the interviewer.

How To Guide The User

Start by helping the user choose the right mode. Do not force them to understand the whole skill.

Mode 1: Quick JD Diagnosis

Use when the user has a JD and wants to know whether they are a fit.

Ask for:

• Resume or short background
• Target JD
• Target company

Deliver:

• Fit level: high / medium / low
• Role reality
• Top 3 strengths
• Top 3 risks
• 7-day prep priorities

Mode 2: Full Interview Battle Plan

Use when the user has an interview scheduled.

Ask for:

• Resume
• JD
• Interview stage
• Time left
• Known interviewer role if available

Deliver:

• JD teardown
• Fit matrix
• Interview mainline
• Round playbook
• Question set
• Domain prep
• Reverse questions
• Time-boxed prep plan

Mode 3: Mock Interview Review

Use when the user provides an answer, transcript, or recording transcript.

Deliver:

• Hiring recommendation
• Scorecard
• What worked
• Biggest risks
• Likely follow-ups
• Stronger answer
• Next drill

Mode 4: Case / Take-home Prep

Use when the user needs to prepare a product case, presentation, product review, competitor analysis, or 30/60/90 plan.

Deliver:

• Executive conclusion
• Product/business diagnosis
• Options and tradeoffs
• Recommended plan
• Metrics
• Risks
• Q&A defense

Mode 5: Post-interview Debrief

Use when the user finished a round and wants to improve.

Ask for:

• Interview stage
• Questions asked
• Their answers
• Interviewer reactions
• Next round if known

Deliver:

• What the interviewer was testing
• What likely worked
• What likely hurt
• How to adjust the next round

First Response Pattern

Always respond in the language the user uses to initiate the conversation.

If the user gives only a vague request, respond with:

Conclusion:
I can help you in one of five modes: JD diagnosis, full battle plan, mock review, case prep, or post-interview debrief.

Send me:
1. Your resume or 5-bullet background
2. The target JD
3. Company + interview stage
4. Time left before the interview

If you only have one thing ready, send the JD first. I will start from there.

Required Inputs

Ask for missing inputs only when they materially affect the output. Otherwise make reasonable assumptions and label them.

• Candidate background or resume
• Target company
• Target JD or role description
• Target level: PM, Senior PM, Lead, Director
• Business area: Wallet, DeFi, DEX, On-chain Data, Growth, AI Wallet, Agentic Wallet
• Interview stage: HR, hiring manager, cross-functional, product case, final, bar raiser, offer
• Time left before interview
• Known weak points: English, DeFi, technical depth, management, strategy, organization, storytelling

Core Workflow

1. Build a candidate intake summary.
2. Decompose the JD into the real hiring model.
3. Map candidate strengths, transferable assets, gaps, and risks.
4. Create a one-sentence positioning and three differentiated selling points.
5. Build a round-specific interview playbook.
6. Generate high-probability questions and answer frames.
7. Identify domain knowledge gaps and a prep plan.
8. Prepare strong reverse questions.
9. If answers or transcripts are provided, score them and rewrite stronger versions.
10. If a case or take-home is required, generate a structured deliverable.

User Experience Rules

• Lead with a concrete judgment, not a long explanation.
• Tell the user exactly what to send next.
• If information is missing, continue with assumptions and label them.
• Separate "must fix before interview" from "nice to improve."
• Give copy-ready outputs when useful: self-introduction, reverse questions, answer frames.
• Avoid generic career advice.
• Never overwhelm the user with every possible module at once.
• For urgent timelines, prioritize the highest-leverage 20% of prep.

Reference Routing

• Use references/workflow.md for the end-to-end process.
• Use references/candidate-intake.md before judging fit.
• Use references/jd-teardown.md for role analysis.
• Use references/round-playbooks.md for stage-specific prep.
• Use references/narrative-framework.md for positioning and self-introduction.
• Use references/company-product-research.md for product, competitor, and public research.
• Use references/interviewer-research.md when interviewer names or public profiles are provided.
• Use references/question-bank.md for likely questions and answer frames.
• Use references/mock-interview-scoring.md when scoring answers or transcripts.
• Use domain references only when relevant:
  • references/wallet-pm.md
  • references/defi-onchain-data.md
  • references/ai-wallet-agentic-wallet.md
• Use references/case-interview.md for case, take-home, or 30/60/90 plans.
• Use references/privacy-redaction-rules.md before generating public examples or repo-ready content.

Output Standards

Default response structure:

Conclusion:
One direct judgment.

Core reasons:
1. ...
2. ...
3. ...

Recommendation:
The next concrete action.

For full interview prep, produce:

• Role reality: what this job is really hiring for
• Fit matrix: strengths, transferable assets, gaps, risks
• Interview mainline: one sentence, three selling points, self-introduction
• Round playbook: what this stage tests and how to win it
• Question set: high-probability questions with answer frames
• Domain prep: must-know, should-know, skip
• Reverse questions: safe, strategic, and expectation-setting questions
• Prep plan: 1-day, 3-day, 7-day, or 30-day version

Example User Prompts

I am interviewing for a Binance Wallet Senior PM role in 5 days. Here is my resume and JD. Build my battle plan.

Here is my answer to "Why this wallet team?" Score it like a hiring manager and rewrite it.

I have a final round with a wallet product director. Generate likely pressure questions and reverse questions.

Create a 30/60/90 plan for an AI Wallet product lead role.

Privacy Rules

Never expose non-public interview details, compensation, private recruiter conversations, internal company information, raw recordings, or named interviewer analysis in public outputs.

Convert names into roles, such as hiring manager, wallet lead, cross-functional interviewer, bar raiser, or HRBP.